IP Subnetting: Practical Subnet Design and Address ...

1 The TCP/IP Guide - Version (Contents)` 382 _ 2001-2004 Charles M. Kozierok. All Rights subnetting : Practical Subnet Design and Address Determination ExampleWhen educators ask students what they consider to be the most confusing aspect in learning about networking, many say that it is IP Address subnetting . While subnetting isn't all that difficult in concept, it can be a bit mind-boggling in part due to the manipulations of binary numbers required. Many people understand the ideas behind subnetting but find it hard to follow the actual steps required to Subnet a this reason, even though I explained the concepts behind subnetting in detail in the previous section, I felt it would be valuable to have another section that provides a step-by-step look at how to perform custom subnetting .

2 This section divides subnetting into five relatively straight-forward stages that cover determining requirements, making the Design decision of how many bits to use for Subnet ID and host ID, and then determining important numbers such as the Subnet mask, Subnet addresses and host focus in this section is on showing the Practical how of subnetting . The topics here work through two examples using a Class B and a Class C sample network to show you how subnetting is done, and I am explicit in showing how everything is calculated.

3 This means the section is a bit number-heavy . Also, I try not to duplicate conceptual issues covered in the previous section, though a certain amount of overlap does occur. Overall, if you are not familiar with how subnetting works at all, you will want to read that section first. I do refer to topics in that section where appropriate, especially the summary tables. Incidentally, I only cover conventional subnetting here, not section may serve as a useful refresher or summary of subnetting for someone who is already familiar with the basics but just wants to review the steps performed in subnetting .

4 Again, always bear in mind that subnetting is based on the older classful IP addressing scheme, and today's Internet is classless, using CIDR. Background Information: If you are not familiar with binary numbers, binary-to-decimal conversion and masking, and you didn't take my advice in preceding sections to brush up on these concepts using the background explanation of computational math, you really want to do that : If in reading this section you find yourself wanting to do binary-to-decimal conversions or binary math, remember that most versions of Windows (and many other operating systems) have a calculator program that incorporates scientific TCP/IP Guide - Version (Contents)` 383 _ 2001-2004 Charles M.

5 Kozierok. All Rights subnetting Step #1: Requirements AnalysisWhen you are building or upgrading a network as a whole, the first step isn't buying hardware, or figuring out protocols, or even Design . It's requirements analysis, the process of determining what it is the network needs to do. Without this foundation, you risk imple-menting a network that may perfectly match your Design but not meet the needs of your organization. The exact same rule applies to subnetting as well. Before we look at the gory details of host addresses and Subnet masks, we must decide how to Subnet the network.

6 To do that, we must understand the requirements of the subnetting RequirementsAnalyzing the requirements of the network for subnetting isn't difficult, because there are only a few issues that we need to consider. Since requirements analysis is usually done by asking questions, here's a list of the most important questions in analyzing subnetting requirements: What class is our IP Address block? How many physical subnets are on the network today? (A physical Subnet generally refers to a broadcast domain on a LAN; a set of hosts on a physical network bounded by routers.)

7 Do we anticipate adding any more physical networks in the near future, and if so, how many? How many hosts do we have in the largest of our subnets today? How many hosts do we anticipate having in the largest Subnet in the near future?The first question is important because everything in subnetting is based around dividing up a Class A, Class B or Class C network, so we need to know which we are dealing with. If we are in the process of designing a network from scratch and don't have a Class A, B or C block yet, then we will determine which we need based on the approximate size of the organization.

8 After that, we need to determine two key numbers: how many physical subnets we have, and the maximum number of hosts per Subnet . Assessing Future Needs During Requirements AnalysisWe need to analyze the requirements above not only for the present network, but for the near future as well. The current values for these two numbers represent how the network needs to be designed today. However, designing only for the present is not a good we have exactly four subnetworks in our network now. In theory, we could use only two bits for the Subnet ID, since 22 is 4.

9 However, if our company is growing rapidly, this would be a poor choice. When we need to add a fifth Subnet we'd have a problem!Similarly, consider the growth in the number of hosts in a Subnet . If the current largest Subnet has 60 hosts, you don't want 6a bits for the host ID, because that limits you to 62 hosts. You can divide large subnets into smaller ones, but this may just mean unnecessarily additional TCP/IP Guide - Version (Contents)` 384 _ 2001-2004 Charles M. Kozierok. All Rights , what is the near future ?

10 The term is necessarily vague, because it depends on how far into the future the organization wants to look. On the one hand, planning for several years' growth can make sense, if you have enough IP addresses to do it. On the other, you don't want to plan too far out, since changes in the short term may cause you to completely redesign your network Concept: To successfully Subnet a network, you must begin by learning what the requirements of the network will be. The most important parameters to determine are the number of subnets required and the maximum number of hosts needed per Subnet .